Three phases in the development of depression have been identified: (1) low prevalence in childhood (1- 2%), (2) low overall prevalence but emergence of a sex difference in early adolescence, and (3) a dramatic rise in prevalence in middle to late adolescence (from 3-17%). It is essential to study vulnerabilities for developing depression in these different phases so that more focused developmentally sensitive treatment / prevention programs may be created for this significant public health concern. This study aims to examine how genetic, cognitive, and interpersonal vulnerability factors work together to contribute to depression within a vulnerability- stress framework during each of these phases. More specifically we seek to study (1) when and how cognitive and interpersonal vulnerabilities stabilize into trait-like risks that interact with stress to predict depression, (2) how genetic, cognitive, and interpersonal vulnerabilities, alone and in interaction with stress, explain why more girls become depressed than boys in early adolescence, and (3) how genetic, cognitive, and interpersonal vulnerabilities relate to each other and interact with stress to predict the dramatic surge in depression in middle to late adolescence. This multi-wave, multi-method (questionnaire, clinical interviews, behavioral tasks, blood draws), multi- informant (youth, mother, clinician) prospective study will (1) explore when psychosocial vulnerabilities to depression stabilize and (2) test the hypotheses that genetic, cognitive, and interpersonal vulnerability factors to depression interact with an increasing number of stressors throughout adolescence, particularly for girls, to predict the emergence of sex differences in the prevalence of depression in early adolescence and the dramatic increase in depression during middle adolescence. To examine causes of depression in each of these phases within this integrative vulnerability-stress framework, we will use an accelerated longitudinal, cross-sequential design to follow 3rd, 6th, and 9th graders and their mothers (n=250 youth and 250 mothers in each cohort), after an initial lab assessment, every 3 months for 3 years (12 follow-up assessments). Multilevel growth curve modeling will be used to test hypotheses.